# Recognition of an obstacle in a flow using artificial neural networks

## TL;DR

This paper develops artificial neural networks to estimate the size and location of obstacles in a pipe flow by analyzing pressure and velocity profiles, demonstrating good predictive capacity with some limitations.

## Contribution

The work introduces neural networks trained on simulation data to accurately estimate obstacle characteristics in fluid flow, a novel approach in flow obstacle detection.

## Key findings

- Neural networks effectively predict obstacle size and position.
- Good estimation accuracy with high R^2 scores.
- Difficulty in classifying very small obstacles.

## Abstract

In this work a series of artificial neural networks (ANNs) have been developed with the capacity to estimate an obstacle's size and location obstructing the flow in a pipe. The ANNs learn the size and location of the obstacle by reading the profiles of the dynamic pressure $q$ or the $x$-component of the velocity $v_x$ of the fluid at certain distance from the obstacle. The data to train the ANN, was generated using numerical simulations with a 2D Lattice Boltzmann code. We analyzed various cases varying both the diameter and position of the obstacle on $y$-axis, obtaining good estimations using the $R^2$ coefficient for the cases of study. Although the ANN showed problems for the classification of the very small obstacles, the general results show a very good capacity of prediction.